- Email: [email protected]
Screening and prevention of stillbirth
Accepted Manuscript Screening and prevention of stillbirth Prof Gordon C.S. Smith, DSc, FMedSci, Professor & Head of Department
PII:
S1521-6934(16)30075-X
DOI:
10.1016/j.bpobgyn.2016.08.002
Reference:
YBEOG 1634
To appear in:
Best Practice & Research Clinical Obstetrics & Gynaecology
Received Date: 4 May 2016 Revised Date:
26 August 2016
Accepted Date: 30 August 2016
Please cite this article as: Smith GCS, Screening and prevention of stillbirth, Best Practice & Research Clinical Obstetrics & Gynaecology (2016), doi: 10.1016/j.bpobgyn.2016.08.002. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
1
ACCEPTED MANUSCRIPT Screening and prevention of stillbirth
Prof Gordon C.S. Smith, DSc, FMedSci. Professor & Head of Department, Obstetrics and Gynaecology, University of Cambridge; NIHR
RI PT
Cambridge Comprehensive Biomedical Research Centre, Cambridge, CB2 0SW, UK.
Correspondence to:
Prof GCS Smith, Department of Obstetrics and Gynaecology, University of Cambridge, Box 223 The Rosie Hospital, Cambridge, CB2 2SW, UK.
AC C
EP
TE D
M AN U
SC
Tel: +44 (0)1223 336871, Fax: +44 (0)1223 215327, E-mail: [email protected]
2
ACCEPTED MANUSCRIPT Abstract (150 words)
Stillbirth is delivery of a baby at or after 24 weeks of gestational age (UK definition) showing no signs of life. It affects almost 1 in 200 pregnancies and is the single major cause of perinatal death. Stillbirth is associated with a wide range of maternal demographic characteristics, but most of the
RI PT
variation in stillbirth risk is independent of these characteristics. Stillbirth is the end point of multiple processes, but the single most common cause is probably placental dysfunction. Stillbirth is
associated with a range of biochemical and ultrasonic predictors, but there is limited evidence to support population based screening. However, the evidence based is weak due by use of poorly characterised screening tests, the failure to couple risk assessment with a clearly effective
SC
intervention for those who screen positive, and inadequate study sample sizes. Basic research needs
M AN U
to identify better predictors, and clinical trials need to adopt more rigorous methodology.
Key words:
AC C
EP
TE D
Stillbirth, screening, intervention, ultrasound, biomarkers.
3
ACCEPTED MANUSCRIPT 1. Introduction The 2016 Lancet Ending Preventable Stillbirth Series indicated that the UK was in about the middle of the range of 49 high income countries in relation to stillbirth rates, and had one of the slowest rates of decline in stillbirth.1 The absolute risk of stillbirth from 24 weeks onwards is 3-5 per 1,000, i.e. about the same as the total risk of death in the first year of life. Moreover, a significant proportion of
RI PT
stillbirths potentially could have been prevented if the babies had been identified as high risk, i.e. cases where the cause of death was not a major congenital anomaly and where the death occurred at a gestational age associated with a low risk of infant mortality.2 A recent study, supported by the James Lind Alliance, reported the top research priorities in relation to stillbirth. This study elicited 1275 responses from 574 participants (equally divided between professionals and non-professionals)
SC
and identified several priorities (out of 300 indicative unanswered questions) directly related to screening, (e.g. Priority #5, Does ultrasound assessment of fetal growth in the third trimester reduce
M AN U
stillbirth?), and 4 others which were also relevant (e.g. Priority #6, Would increasing the frequency of umbilical artery Doppler scanning during pregnancy reduce stillbirth?).3 The top 11 recommendations are listed (Table 1). The aim of the current review is to consider current practice in relation to screening for stillbirth, and how improved methods of screening may be developed and evaluated.
TE D
1.1 Epidemiology of stillbirth
Stillbirth is defined as delivery of a baby in the perinatal period which fails to show any signs of life. The definition of the start of the perinatal period varies between countries. In the USA it is 20 weeks of gestational age (wkGA) and in the UK it is 24wkGA.2 The rate of stillbirth in high income countries
EP
varies from 1.3 per 1,000 to 8.8 per 1,000, with an average of 3.5 per 1,000.1 The clear majority of stillbirths in high income countries are the result of intra-uterine fetal death prior to the onset of
AC C
labour (antepartum stillbirth).2 Deaths occurring during labour (intrapartum stillbirths) account for 510% of all stillbirths in high income countries but account for a much larger proportion of losses in low and middle income countries (LMIC).4 A range of maternal characteristics have been associated with the risk of stillbirth. One of the most systematic and methodologically sound analyses of such risk factors was performed by the NICHD's Stillbirth Collaborative Research network.5 Their analysis of risk factors identifiable at the start of pregnancy is tabulated (Table 2).
1.2 Causes and classification of stillbirth Aside from the sub-division of stillbirths into intrapartum and antepartum, losses can also be classified according to the presumed cause. However, this process is complicated by the fact that
4
ACCEPTED MANUSCRIPT relatively few losses have a completely understood cause of death. The remainder have varying degrees of uncertainty in the mechanism(s) leading to death. This is illustrated in Figure 1 in relation to ascribing cause of death in the presence of a range of maternal medical conditions. A diverse – if not bewildering – array of classification systems have been developed.1 One of the main characteristics which determines variation between the systems is the extent to which they
RI PT
will attribute a given associated condition or finding as being the cause of death. For example, an unexpected and unexplained stillbirth of an infant at 39 weeks where the baby's birth weight was on the 2rd percentile for sex and gestational age might be defined as unexplained in one classification system and as being due to fetal growth restriction (FGR) in another. In a sense, both classifications are correct. The actual cause of death is unknown, hence the loss is strictly unexplained. However, it
SC
is known that birth weight <3rd percentile is associated with a 10-fold risk of stillbirth at term.6
Hence, it is very likely that the baby's death was related to its size. The subject of classification of
M AN U
stillbirth is reviewed in detail elsewhere.7
In the context of screening, one of the key associations for stillbirth is poor fetal growth. It is estimated that between 30 and 50% of stillbirths are associated with low birth weight percentile, and this is assumed to reflect FGR, which is in turn presumed to be related, in a large proportion of cases, to placental dysfunction.2 This association is important in the context of screening, as a
2. Screening for stillbirth
TE D
number of tools exist to quantify both fetal growth and placental function.
The first element of screening is to differentiate a population into those at high risk of the condition and those at low risk. The analysis and interpretation of screening statistics can be complicated.
EP
Particular problems in the context of stillbirth are as follows. (i) stillbirth is the end result of diverse pathological processes. Hence, no one test is likely to be
AC C
highly sensitive for the condition overall. (ii) stillbirth can occur across the whole range of gestational age. As the primary means to prevent stillbirth is to deliver the baby, the consequences of inappropriate intervention differ profoundly: at 24 weeks outcomes are generally extremely poor, whereas at 40 weeks, outcomes are generally extremely good.
(iii) stillbirth is relatively uncommon, hence, any sub-type of stillbirth is less common still. For example, if 40% of stillbirths are due to FGR and the risk of stillbirth in a country is 4 per 1,000, the absolute risk of stillbirth associated with FGR is 1.6 per 1,000. Normally, a test with a positive likelihood ratio of 20 would be regarded as an excellent screening test. However, the positive predictive value of this test in such a population would be about 3%. Consequently, 97% of screen
5
ACCEPTED MANUSCRIPT positive women would not experience the event of interest. It follows, therefore, that if an intervention based on such a test caused any harm to false positives, that a programme of screening and intervention would be likely to cause harm despite a strong screening test and an effective intervention.
RI PT
3. Maternal risk factors The maternal characteristics which are associated with the risk of stillbirth have been tabulated (Table 2). However, collectively, these associations explained just 19% of the variability in the risk of stillbirth. Hence, programmes of screening and intervention which focus on maternal risk factors will have a relatively limited capacity to reduce the numbers of stillbirth. Another key element of
SC
maternal risk are the associations with gestational diabetes mellitus (GDM). There are a series of maternal characteristics which place a woman at increased risk of GDM (Table 3). A recent detailed
M AN U
review of a representative sample of unexpected stillbirths of normally formed infants at term in the UK, conducted by the national perinatal mortality surveillance system, MBRRACE-UK, found that failure to screen for GDM was one of the three most common potentially preventable characteristics observed which might have resulted in the loss being avoided.8
3.1 Clinical factors associated with stillbirth risk
TE D
Reduced fetal movements (RFM) is one of the key symptoms associated with the risk of stillbirth. This subject has been reviewed in detail elsewhere.9 Failure of medical and midwifery staff to respond appropriately to RFM was one of the other major potentially preventable causes of stillbirth in the MBRRACE-UK review described above.8 The RCOG in the UK has issued a guideline for the
EP
management of women who present with reduced movements.10 Another major symptom associated with stillbirth risk is antepartum haemorrhage. In the absence of placenta praevia or
AC C
cervical pathology, this is likely to represent bleeding from the placenta, and may be a harbinger of later abruption. Finally, acquired disorders of pregnancy (such as GDM or pre-eclampsia) could be associated with maternal symptoms. Management is based on specific testing for the condition in question.
3.2 Response to women with symptoms associated with stillbirth risk At preterm gestational ages, fetal monitoring is indicated. Generally, the first line of assessment is cardiotocography (CTG, also called a non-stress test in the USA). Interestingly, the RCOG Guideline on RFM does not recommend computerised CTG,10 which utilises objective analysis of the trace, including computerised assessment of beat to beat variability. However, the Cochrane reviews
6
ACCEPTED MANUSCRIPT indicate no benefit of using non-computerised CTG (and a trend towards harm) compared with using nothing, and a reduced risk of perinatal death using computerised CTG analysis compared with noncomputerised CTG (see Smith 2015, for review).11 Hence, computerised CTG is a reasonable first step. The next level of assessment is an ultrasound scan. This is done if there are other risk factors present, but will also be done in low risk women who are presenting with their second (or greater)
RI PT
episode of RFM. The key evidence-based measurement in high risk pregnancies is Doppler flow velocimetry of the umbilical artery. However, generally, fetal biometry and amniotic fluid
measurement would also be performed, sometimes with additional Doppler measurements (middle cerebral artery). The evidence base supporting these supplementary measurements is poor: this is a key area for further research and multiple recommendations of the prioritisation exercise for
SC
stillbirth research touched on this area.
M AN U
At term, induction of labour should be considered for women presenting with RFM. At 37-38 weeks, induction is still associated with increased risks of perinatal morbidity and should only be offered if fetal assessment is non-reassuring, there are other risk factors or if there have been multiple episodes of RFM. However, perinatal outcome is optimal at 39-41 weeks and induction should be considered for any women presenting with RFM at ≥39 weeks. Epidemiological evidence exists that universal induction at 39 weeks would reduce overall rates of stillbirth.12 However, for women with a
TE D
single episode of RFM, who have no other risk factors, and who want to avoid intervention, it is reasonable not to induce labour, as the absolute risk of stillbirth is likely to be <1%. The basis for this statement is that the background risk of stillbirth at term is 1-2 per 1,000 and it is unlikely that a
EP
single episode of RFM in an otherwise low risk women is associated with a >5-fold risk of stillbirth.
4. Biochemical predictors of stillbirth risk
AC C
A range of biochemical tests have been associated with the risk of stillbirth. Most of these are maternal blood tests but, historically, urine tests were also evaluated. Many of the associations described were secondary analyses of measurements made for the purposes of screening for unrelated conditions, such as neural tube defects and Down's syndrome.
4.1 Pregnancy associated plasma protein A (PAPP-A) PAPP-A is a protease for insulin like growth factor binding proteins 4 and 5, and it is produced by the placenta. Its primary application has been for screening for Down's syndrome. However, it was shown that low first trimester PAPP-A was associated with stillbirth risk and that the association was due to losses related to placental dysfunction.13;14 A systematic review concluded that PAPP-A was a
7
ACCEPTED MANUSCRIPT "good predictor of stillbirth related to placental dysfunction disorders".15 In the UK, an RCOG Guideline recommends methods for further monitoring of women found to have low PAPP-A in the first trimester of pregnancy.16
4.2 Alpha fetoprotein (AFP)
RI PT
AFP is a fetal oncotic protein. High maternal serum levels are associated with structural fetal anomalies (neural tube defects and abdominal wall defects) and low levels are associated with
Down's syndrome. High levels in the second trimester are associated with the risk of stillbirth in normally-formed babies, but the associations are stronger for losses occurring at extreme preterm gestational ages.17 Hence, intervention to prevent such losses is problematic. The association is
SC
thought to be related to abnormal permeability of the placenta, hence increased passage of the protein across the placental barrier. Interestingly, unexplained raised maternal serum AFP is
where the pregnancy was complicated.18
4.3 Others
M AN U
associated with a diverse series of abnormalities on subsequent placental pathological examination
A range of other biochemical tests have been associated with the risk of FGR and stillbirth. Some of these were in widespread use prior to the advent of ultrasound, such as measurement of oestrogens
TE D
in maternal urine. Others have been more recently identified as potential screening tests for preeclampsia, in particular pro- and anti-angiongenic proteins circulating in the mother’s blood, such as PlGF, sFlt-1, and eEng. Studies have described some promising associations with the angiogenic regulators.19 However, the relative rarity of stillbirth (compared with preeclampsia and preterm
EP
birth) make it a more difficult subject to address. Cohort studies have to be larger in order to be able to detect associations. Moreover, if a measurement has a positive likelihood ratio of 10, which
AC C
would usually be regarded as a good screening test, it would only have a positive predictive value of <5%, and much lower still if performed later in pregnancy.
Biochemical tests have great potential as a screening tool for stillbirth when combined with ultrasound. However, any such potential will not be adequately evaluated unless studies are sufficiently powered to detect associations. Moreover, interventional studies will present huge challenges and will require addressing significant methodological considerations (see below).
5. Ultrasonic assessment of stillbirth risk
8
ACCEPTED MANUSCRIPT The first use of ultrasound to estimate the weight of the fetus was described in 1975.20 Subsequently, multiple other methods were described to assess fetal well-being, including biophysical measurements (activity, breathing movements, tone and liquor volume) and ultrasonic Doppler flow velocimetry of fetal and utero-placental arteries and veins. Research focused on both the diagnosis of fetal compromise in women who had presented with complications and/or those
RI PT
known to have risk factors, and also on trying to identify which apparently low risk women had occult fetal compromise. The body of work quickly led to RCTs of universal screening of low risk women using late pregnancy ultrasound. A meta-analysis of these trials (latest version is Bricker et al 201521 including data from a total of 34,980 women recruited to 13 trials), reported no maternal or fetal benefit of screening. A previous version of this meta-analysis (which had very similar
M AN U
not have routine late pregnancy ultrasound scans.
SC
conclusions) led NICE to conclude in their 2008 Guideline on Antenatal Care22 that women should
In the same Guideline, NICE indicated that further studies on the diagnostic effectiveness of universal ultrasound as a screening method in the general population should be considered a priority. It may seem counterintuitive that NICE would declare universal screening to be ineffective but then to propose further studies on the diagnostic effectiveness of the screening test. This
TE D
apparent anomaly is explained by closer examination of the studies included in the meta-analysis. Ultimately, all meta-analyses, however appropriately conducted, are only as good as the trials included. Issues with the 13 trials in the Cochrane review are as follows: (i) the individual trials used different definitions of screen positive, (ii) none of the definitions had been evaluated in a level 1
EP
study of diagnostic effectiveness prior to being employed in the screening study, i.e. all of the individual interventional studies were designed and conducted in the absence of high quality
AC C
quantitative information about the performance of the screening test, (iii) none of the trials had a standardised intervention for those who screened positive, other than to reveal the test or to perform further scans. Finally, even the meta-analysis itself is under-powered. The 2016 Lancet Ending Preventable Stillbirth Series reported sample size calculations for trials of screening and intervention for stillbirth. These looked at the statistical power assuming very optimistic values of screening test performance (5% of the population identified with a positive likelihood ratio of 10) and intervention (50% reduction) based on the incidence of stillbirth (1 in 200). The sample size required (90% power, alpha 0.05, two sided) was 131,000.1 Hence, the meta-analysis has only about a quarter of the sample size required for even a very good screening test coupled with a highly effective intervention. Consequently, despite the fact that ultrasound has been used for >40 years to assess fetal well-being, and despite the fact that it is the primary method for assessing and
9
ACCEPTED MANUSCRIPT monitoring high risk pregnancies, there is an absence of evidence in relation to its utility in screening low risk women. Universal ultrasound has been assessed as a means of detecting small for gestational age (SGA) infants in a level 1 study of diagnostic effectiveness (i.e. where the result of the research scans was blinded) including ~4,000 first pregnancies.23 The rate of detection of SGA infants was 20% with
RI PT
selective use of ultrasound and 57% with universal scanning. However, for every one additional true positive result, there were two additional false positives with universal scanning. Moreover, of a range of ultrasonic indicators of FGR, only one differentiated between SGA infants and the risk of perinatal morbidity. Fetuses which were SGA on scan and had reduced growth velocity of the
SC
abdominal circumference (compared with the 20 week scan) were at increased risk of morbidity compared with normal sized fetuses, whereas SGA was not associated with the risk of morbidity if the growth velocity was normal. A meta-analysis has also demonstrated that high resistance
M AN U
patterns of Doppler flow velocimetry in the uterine arteries in mid-gestation are a good predictor of stillbirth related to placental dysfunction.15 However, the studies revealed the result of the scan, which complicates assessment of the association.
Many private ultrasound providers have started offering late pregnancy ultrasound scans. There is, therefore, ad hoc screening at present with very little evidence regarding the balance of risks and
TE D
benefits. Late pregnancy ultrasound has the potential to cause harm. False positive diagnoses can lead to unnecessary intervention and, in particular, early term delivery. While this is a valuable and evidence based intervention in certain contexts, such as pre-eclampsia, infants born at 37-38 weeks are at increased risk of short term morbidity,24 and even have poorer achievement at school.25
EP
Hence, ad hoc screening and intervention has the potential to cause short and long term harm through iatrogenic late preterm and early term birth. This has recently been recognised in France,
AC C
where routine ultrasound was implemented in the absence of supportive evidence. Screening did not appear to confer any benefit, but false positive diagnosis of SGA was associated with a greater than 3-fold increase in the risk of medically-indicated preterm birth.26 This does not indicate that universal screening is futile. However, these findings underline the importance of understanding what elements of prenatal ultrasonic assessment of the fetus in a low risk pregnancy are informative of risk, and what interventions can be applied to mitigate the risk without causing harm.
6. Interventions to prevent stillbirth 6.1 Medical interventions 6.1.1 Aspirin
10
ACCEPTED MANUSCRIPT Low dose aspirin has been widely evaluated as a method for preventing placentally-related complications in pregnancy and, in particular, pre-eclampsia. Aspirin acts by irreversibly inhibiting the enzyme cyclo-oxygenase (COX), which is a key point in the biochemical synthesis of prostanoids, including thromboxane. Aspirin selectively blocks platelet prostanoid production as they lack a nucleus and cannot synthesise further COX following the irreversible binding with aspirin. The Cochrane review contains 14 trials containing 33,098 women, and demonstrates an 14% reduction in
RI PT
the risk of stillbirth or neonatal death.27 Hence, aspirin should be considered for all women at risk of stillbirth. Typically, it would be started at 12 weeks and discontinued at 36 weeks. There is
preliminary evidence from the pre-eclampsia literature to suggest that starting the treatment prior to 16 weeks may be associated with greater risk reduction.28
SC
6.1.2 Low molecular weight heparin
Given the success of aspirin and given some published associations between maternal thrombophilia
M AN U
and stillbirth risk, many groups have considered the use of LMWH as a preventative treatment for women at increased risk of stillbirth. It is regarded as a strongly indicated therapy in some situations, such as anti-phospholipid syndrome, where it has the additional benefit of reducing the risk of maternal venous thrombo-embolism. A series of small trials appeared to show a benefit of the treatment on perinatal outcome.29 However, a subsequent multicentre study demonstrated no beneficial effect of LMWH.30 A caveat to this is that even that study was only powered to detect very
TE D
major effects of LMWH. The confidence intervals around the effect from the trial include reductions in stillbirth risk which would be regarded as clinically significant. The current status of LMWH is, therefore, that it’s use should primarily be considered to prevent maternal venous thromboembolism and that further studies are required before its use is recommended for the primary aim
EP
of preventing fetal complications.
AC C
6.1.3 Nitric oxide donors and selective phosphodiesterase inhibitors Nitric oxide, a gaseous mediator of smooth muscle relaxation, is thought to have a key role in the control of placental development, with a deficit of nitric oxide effects postulated to be a determinant of abnormal placentation.31 This has led to studies evaluating methods for increasing nitric oxide. Broadly, these fall into two categories: (i) nitric oxide donors, and (ii) selective phosphodiesterase inhibitors. Many nitric oxide donors have been developed for treatment of cardiovascular disease, such as angina and heart failure. A number of studies have evaluated nitric oxide donors in high risk populations. These have generally been small scale single centre studies and some have generated positive results.32;33 However, further large scale multicentre trials are required prior to introduction of these agents in the management of high risk women. Selective
11
ACCEPTED MANUSCRIPT phosphodiesterase inhibitors were first developed for cardiovascular applications. But early experience in healthy volunteers led to an unanticipated effect on male erectile function, and a range of these drugs is available for erectile dysfunction. The best known is sildenafil citrate (known commercially as Viagara), and the evidence supporting a beneficial effect of this drug in the context of early onset FGR has been reviewed.34 Currently, a multicentre RCT is in process, STRIDER, which
RI PT
seeks to assess the effect of sildenafil citrate in pregnancies complicated by early onset severe FGR. The field may also develop into the evaluation of other selective PDE inhibitors with longer half-life, and the use of the drug in pregnancies at high risk of FGR by their screening results, but prior to the disease onset.
SC
6.1.4 Others
A range of other specific therapies exist, which are at various stages of evaluation but are not likely
M AN U
to be in routine clinical use in the very near future. These include the use of supplemental oxygen35 and gene therapy.36 Both of these approaches face significant logistical issues. Gene therapy has been evaluated in animal models and is currently being assessed for acceptability and feasibility in the context of severe, early onset FGR.
Finally, optimal management of co-existing medical conditions is an indirect way to reduce the risk
TE D
of stillbirth associated with maternal disease. The obvious example is diabetes mellitus, both gestational and pre-existing. The risk of perinatal mortality reduces with better control. Although diabetes is the obvious example, it is plausible that optimal medical therapy would improve perinatal
disorders etc.
EP
outcome with other maternal medical conditions, such as thyroid disease, connective tissue
AC C
6.2 Delivery to prevent stillbirth
At present, the only intervention which has a major effect on the risk of stillbirth is delivery. The effect of this clearly depends on (i) the background risk of stillbirth, and (ii) the gestational age, as GA is the major determinant of the risk of neonatal death. At term, the risk of neonatal death is extremely low, particularly following exclusion of deaths due to anomalies. However, the weekly risk of stillbirth is stable from 24 onwards and then rises at term and postterm. Combining analysis of declining risk of neonatal death and increasing and cumulative risk of stillbirth, it has been estimated that the risk of perinatal death is lowest at around 39 weeks.12 Hence, one approach to preventing term stillbirths would simply be to induce labour in all women at 39 weeks. This approach is supported by a Cochrane review, which indicates that routine induction of labour at term and post
12
ACCEPTED MANUSCRIPT term reduces the risk of perinatal mortality by ~70%.37 However, most would feel that this level of intervention would be excessive in relation to the number of deaths prevented. Nevertheless, this background should inform attitudes in general. If a woman feels strongly that she should be induced at 39 weeks or later, it is probably not unreasonable to agree to the request assuming the healthcare system is sufficiently well resourced. Furthermore, a low threshold for induction of labour should be
RI PT
applied to women with concerning symptoms or investigations. If a woman presents with reduced fetal movements (RFM) at 34 weeks, induction of labour would only be considered if there were confirmatory tests indicating other concerns. However, at 39 weeks and beyond, it would be entirely reasonable to induce labour for RFM. This would be especially the case if the woman actively wanted induced, and it would be appropriate if the woman was neutral and expressed the sentiment of so
SC
many “I just want what is best for my baby”. However, many women are keen to avoid induction. In such cases, an open discussion is required. Women who decline induction can be reassured that,
M AN U
with a single episode of RFM, no associated risk factors, and reassuring monitoring, the absolute risk of a stillbirth remains low. It is essential in all these discussions that the woman’s own attitudes and preferences are a foundation for the discussion. A key element for any such discussion would be to contradict the misconception that induction of labour increases the risk of caesarean section. A systematic review indicates that it results in a lower risk of caesarean delivery.38 Moreover, a recent RCT has shown no increase in the risk of caesarean section when labour was induced at 39 weeks in
greater).39
TE D
a group of women with a high background risk of caesarean (nulliparous women aged 35y or
6.3 Modifying maternal risk factors for stillbirth
EP
A number of the maternal characteristics associated with stillbirth risk are potentially modifiable. Numerous studies have demonstrated higher rates of stillbirth among smokers, and a causal
AC C
association is plausible. It follows, therefore, that smoking cessation is likely to reduce the risk of stillbirth. Given the problems of conducting RCTs of interventions to prevent stillbirth (see below), it is unlikely that there will ever be level 1 evidence to support the use of smoking cessation as a means to prevent stillbirth. Nevertheless, given the background and given that this represents an absence of evidence, rather than evidence of absence, recommending that mothers participate in smoking cessation would be an appropriate approach. Similarly, raised body mass index is both directly associated with stillbirth in epidemiological studies and is also a risk factor for a cause of stillbirth, namely, GDM. Weight reduction should be considered for women who are obese and are planning pregnancy. Dieting through calorific restriction is controversial in pregnancy. However, women who are obese should be referred to a dietician and approaches which might improve
13
ACCEPTED MANUSCRIPT outcomes include reducing quantities of sugar intake, controlling portion sizes, eating more fresh fruit and vegetables, and the maintenance of healthy levels of activity through pregnancy. A number of epidemiological studies have shown associations between maternal sleep position and the risk of stillbirth. These have demonstrated increased risks of stillbirth with supine sleep position, and it is speculated that this may be due to reduced blood supply to the uterus through caval
RI PT
compression (i.e. a pressure effect of the gravid uterus on the inferior vena cava). Further studies aiming to confirm or refute the association are in progress. If these confirm the association with maternal sleep position, there may be a case for a public health campaign to modify maternal sleep position as a means to reduce stillbirth risk. This is analogous with sudden infant death syndrome: epidemiological studies demonstrated an association with prone infant sleeping positon and SIDS,
SC
and a public health campaign was followed by an extraordinarily rapid decline in rates of SIDS in the
M AN U
UK.40
7. Evaluating novel methods for screening and prevention of stillbirth. Despite huge expenditure of time and resources, there is no clear approach to screening women for stillbirth risk. Numerous studies have described possible methods for screening, including biochemical tests and ultrasound scans. However, none has been clearly shown to be effective in randomised controlled trials. The negative results of the trials do not necessarily indicate that
TE D
ultrasound is futile. As discussed above, there are multiple issues in relation to the meta-analysis of RCTs of universal ultrasound. It is essential, therefore, that any future trials learn the lessons from previous trials and address new methods of screening using a methodologically rigorous approach. Key elements in designing new RCTs to prevent stillbirth are as follows: (i) the eligible population, (ii)
EP
the screening test to be employed and its diagnostic effectiveness, (iii) the timing of randomisation, namely, before or after application of the screening test, (iv) the intervention, (v) the control group,
AC C
(vi) the primary and secondary outcomes, (vii) the sample size, (viii) the possibility of randomising at the level of hospitals, e.g. using a cluster RCT or stepped wedge RCT, (ix) the potential use of an adaptive trial design, (x) whether an initial smaller scale feasibility study is required, and (xi) the feasibility of the study, including manpower, skills mix, and quality assurance of the screening test. Many of these issues have previously been discussed in more depth elsewhere.41
8. Summary Stillbirth in the single major determinant of perinatal death. In high income countries, ≥90% follow death of the baby prior to the onset of labour, and in this setting typical absolute rates are 3-4 per 1,000. Stillbirth is the end result of diverse processes, hence a single perfect screening test is unlikely
14
ACCEPTED MANUSCRIPT to be developed. Currently, women are assessed for their risk of stillbirth at the time of first attendance for antenatal care, and then are re-assessed during the pregnancy either at routine visits or when they present with symptoms which are associated with increased risk of stillbirth (such as reduced fetal movements). A range of methods have been developed to estimate the risk of stillbirth and associated conditions (such as FGR). Currently, investigations such as ultrasound are only
RI PT
applied to women who have risk factors or attend with complications. However, the possible utility of universal ultrasound has been poorly evaluated. About one third of all stillbirths occur at term and these are, potentially, the most readily preventable as there is a safe and effective intervention, namely, induction of labour, which reduces the risk of perinatal death by 70%. A key challenge is how to identify the women at increased risk of stillbirth and who would benefit from this
M AN U
causing unacceptably high rates of intervention.
SC
intervention, from the general healthy population. In this way, losses could be prevented without
Conflicts of interest.
GS receives/has received research support from GE, Roche and GSK. GS has been paid to attend advisory boards by GSK and Roche. GS has acted as a paid consultant to GSK. GS has received support to attend a scientific meeting from Chiesi. GS is named inventor in a patent submitted by
TE D
GSK (UK), for novel application of an existing GSK compound for the prevention of preterm birth (PCT/EP2014/062602). GS has acted as an expert witness. GS is a member of a Data Safety
Practice points •
EP
Monitoring Committee for a trial of an RSV vaccine in pregnancy, being run by GSK.
Maternal risk factors do not perform well as a screening test for stillbirth. Many stillbirths
•
AC C
occur to low risk women
Common issues with quality of care in relation to stillbirth prevention include failure of
effective screening or treatment of GDM, reduced fetal movements, or poor fetal growth
•
Non-computerised antenatal cardiotocography has little role in assessing fetal well being,
and may provide false reassurance about stillbirth risk in women presenting with reduced movements •
Induction of labour can safely be considered from 39 weeks onwards, even in relatively low risk women. It will prevent subsequent stillbirth, and is not associated with an increased risk of caesarean delivery.
15
ACCEPTED MANUSCRIPT •
Women at risk of stillbirth should be treated with low dose aspirin from 12 to 36 weeks. Low molecular weight heparin should primarily be used to prevent maternal thrombo-embolism and its use to prevent stillbirth will largely be confined to randomised controlled trials
• Research agenda Design, feasibility and optimal design of an RCT of universal ultrasound, and the effect of
RI PT
•
universal ultrasound scanning •
The best ultrasonic markers of likely intra-uterine fetal death in women presenting with complications
•
The ability of biomarkers to differentiate between women with normal and abnormal
•
SC
placental function
The relationship between placental pathology and adverse pregnancy outcome, and
M AN U
methods for diagnosing specific placental pathological processes in the antenatal period (e.g. blood tests for recognised placental disease processes)
EP
TE D
The mechanisms leading to placental dysfunction
AC C
•
16
ACCEPTED MANUSCRIPT Legend for Figure.
AC C
EP
TE D
M AN U
SC
RI PT
Figure 1. Representation of the continuum of uncertainty in ascribing cause for stillbirth, using the example of losses associated with maternal medical conditions. ALT denotes alanine transaminase; GA denotes gestational age. Reproduced with permission from a report of an NICHD workshop (Reddy et al., Obstet Gynecol 2009;114:901-914).
17
ACCEPTED MANUSCRIPT References
*(1) Flenady V, Wojcieszek AM, Middleton P, Ellwood D, Erwich JJ, Coory M et al. Stillbirths: recall to action in high-income countries. Lancet 2016; 387(10019):691-702. *(2) Smith GC, Fretts RC. Stillbirth. Lancet 2007; 370(9600):1715-1725.
RI PT
(3) Heazell AE, Whitworth MK, Whitcombe J, Glover SW, Bevan C, Brewin J et al. Research priorities for stillbirth: process overview and results from UK Stillbirth Priority Setting Partnership. Ultrasound Obstet Gynecol 2015; 46(6):641-647.
SC
(4) Lawn JE, Blencowe H, Pattinson R, Cousens S, Kumar R, Ibiebele I et al. Stillbirths: Where? When? Why? How to make the data count? Lancet 2011; 377(9775):1448-1463. *(5) SCRN Writing Group . Association between stillbirth and risk factors known at pregnancy confirmation. JAMA 2011; 306(22):2469-2479.
M AN U
(6) Moraitis AA, Wood AM, Fleming M, Smith GC. Birth weight percentile and the risk of term perinatal death. Obstet Gynecol 2014; 124(2 Pt 1):274-283. (7) Flenady V, Froen JF, Pinar H, Torabi R, Saastad E, Guyon G et al. An evaluation of classification systems for stillbirth. BMC Pregnancy Childbirth 2009; 9:24.
TE D
(8) Draper ES, Kurinczuk JJ, Kenyon S. MBRRACE-UK Perinatal confidential enquiry: term, singleton, normally formed, antepartum stillbirth. Leicester, UK: The Infant Mortality and Morbidity Studies, Department of Health Sciences, University of Leicester; 2015. (9) Froen JF, Tveit JV, Saastad E, Bordahl PE, Stray-Pedersen B, Heazell AE et al. Management of decreased fetal movements. Semin Perinatol 2008; 32(4):307-311.
EP
(10) RCOG. RCOG Green-top guideline No. 57: Reduced fetal movements. London, UK: RCOG; 2011. (11) Smith GCS. Prevention of stillbirth. The Obstetrician and Gynaecologist 2015;DOI 10.1111/tog.12197.
AC C
(12) Smith GC. Life-table analysis of the risk of perinatal death at term and post term in singleton pregnancies. Am J Obstet Gynecol 2001; 184(3):489-496. (13) Smith GC, Stenhouse EJ, Crossley JA, Aitken DA, Cameron AD, Connor JM. Early pregnancy levels of pregnancy-associated plasma protein a and the risk of intrauterine growth restriction, premature birth, preeclampsia, and stillbirth. J Clin Endocrinol Metab 2002; 87(4):1762-1767. (14) Smith GC, Crossley JA, Aitken DA, Pell JP, Cameron AD, Connor JM et al. First-trimester placentation and the risk of antepartum stillbirth. JAMA 2004; 292(18):2249-2254. *(15) Conde-Agudelo A, Bird S, Kennedy S, Villar J, Papageorghiou A. First- and second-trimester tests to predict stillbirth in unselected pregnant women: a systematic review and metaanalysis. BJOG 2014.
18
ACCEPTED MANUSCRIPT (16) RCOG. Guideline No. 31: The investigation and management of the small for gestational age fetus. 1-34. 2013. London, UK, RCOG Press.
(17) Smith GC, Shah I, White IR, Pell JP, Crossley JA, Dobbie R. Maternal and biochemical predictors of antepartum stillbirth among nulliparous women in relation to gestational age of fetal death. BJOG 2007; 114(6):705-714.
RI PT
(18) Jauniaux E, Moscoso G, Campbell S, Gibb D, Driver M, Nicolaides KH. Correlation of ultrasound and pathologic findings of placental anomalies in pregnancies with elevated maternal serum alpha-fetoprotein. Eur J Obstet Gynecol Reprod Biol 1990; 37(3):219-230.
SC
(19) Chaiworapongsa T, Romero R, Korzeniewski SJ, Kusanovic JP, Soto E, Lam J et al. Maternal plasma concentrations of angiogenic/antiangiogenic factors in the third trimester of pregnancy to identify the patient at risk for stillbirth at or near term and severe late preeclampsia. Am J Obstet Gynecol 2013; 208(4):287.
M AN U
(20) Campbell S, Wilkin D. Ultrasonic measurement of fetal abdomen circumference in the estimation of fetal weight. Br J Obstet Gynaecol 1975; 82(9):689-697. *(21) Bricker L, Medley N, Pratt JJ. Routine ultrasound in late pregnancy (after 24 weeks' gestation). Cochrane Database Syst Rev 2015; 6:CD001451. (22) National Collaborating Centre for Women's and Children's Health. NICE Guideline: Antenatal care. 2008. London, RCOG Press.
TE D
*(23) Sovio U, White IR, Dacey A, Pasupathy D, Smith GCS. Screening for fetal growth restriction with universal third trimester ultrasonography in nulliparous women in the Pregnancy Outcome Prediction (POP) study: a prospective cohort study. Lancet 2015; 386:2089-2097. *(24) Spong CY, Mercer BM, D'alton M, Kilpatrick S, Blackwell S, Saade G. Timing of indicated latepreterm and early-term birth. Obstet Gynecol 2011; 118(2 Pt 1):323-333.
EP
(25) MacKay DF, Smith GC, Dobbie R, Pell JP. Gestational age at delivery and special educational need: retrospective cohort study of 407,503 schoolchildren. PLoS Med 2010; 7(6):e1000289.
AC C
*(26) Monier I, Blondel B, Ego A, Kaminiski M, Goffinet F, Zeitlin J. Poor effectiveness of antenatal detection of fetal growth restriction and consequences for obstetric management and neonatal outcomes: a French national study. BJOG 2015; 122(4):518-527. (27) Duley L, Henderson-Smart DJ, Meher S, King JF. Antiplatelet agents for preventing preeclampsia and its complications. Cochrane Database Syst Rev 2007;(2):CD004659. (28) Bujold E, Roberge S, Lacasse Y, Bureau M, Audibert F, Marcoux S et al. Prevention of preeclampsia and intrauterine growth restriction with aspirin started in early pregnancy: a meta-analysis. Obstet Gynecol 2010; 116(2 Pt 1):402-414. (29) Dodd JM, McLeod A, Windrim RC, Kingdom J. Antithrombotic therapy for improving maternal or infant health outcomes in women considered at risk of placental dysfunction. Cochrane Database Syst Rev 2013; 7:CD006780.
19
ACCEPTED MANUSCRIPT *(30) Rodger MA, Hague WM, Kingdom J, Kahn SR, Karovitch A, Sermer M et al. Antepartum dalteparin versus no antepartum dalteparin for the prevention of pregnancy complications in pregnant women with thrombophilia (TIPPS): a multinational open-label randomised trial. Lancet 2014; 384(9955):1673-1683. (31) Lyall F. Development of the utero-placental circulation: the role of carbon monoxide and nitric oxide in trophoblast invasion and spiral artery transformation. Microsc Res Tech 2003; 60(4):402-411.
RI PT
(32) Schleussner E, Lehmann T, Kahler C, Schneider U, Schlembach D, Groten T. Impact of the nitric oxide-donor pentaerythrityl-tetranitrate on perinatal outcome in risk pregnancies: a prospective, randomized, double-blinded trial. J Perinat Med 2014; 42(4):507-514.
SC
(33) Abdel RM, El-Berry S, Abosereah M, Edris Y, Sharafeldeen A. Prophylactic treatment for preeclampsia in high-risk teenage primigravidae with nitric oxide donors: a pilot study. J Matern Fetal Neonatal Med 2016; 29(16):2617-2620.
M AN U
(34) Panda S, Das A, Md NH. Sildenafil citrate in fetal growth restriction. J Reprod Infertil 2014; 15(3):168-169. (35) Say L, Gulmezoglu AM, Hofmeyr GJ. Maternal oxygen administration for suspected impaired fetal growth. Cochrane Database Syst Rev 2003;(1):CD000137. (36) Spencer RN, Carr DJ, David AL. Treatment of poor placentation and the prevention of associated adverse outcomes--what does the future hold? Prenat Diagn 2014; 34(7):677684.
TE D
(37) Gulmezoglu AM, Crowther CA, Middleton P, Heatley E. Induction of labour for improving birth outcomes for women at or beyond term. Cochrane Database Syst Rev 2012; 6:CD004945.
EP
*(38) Wood S, Cooper S, Ross S. Does induction of labour increase the risk of caesarean section? A systematic review and meta-analysis of trials in women with intact membranes. BJOG 2014; 121(6):674-685.
AC C
(39) Walker KF, Bugg GJ, Macpherson M, McCormick C, Grace N, Wildsmith C et al. Randomized Trial of Labor Induction in Women 35 Years of Age or Older. N Engl J Med 2016; 374(9):813822. (40) Wood AM, Pasupathy D, Pell JP, Fleming M, Smith GC. Trends in socioeconomic inequalities in risk of sudden infant death syndrome, other causes of infant mortality, and stillbirth in Scotland: population based study. BMJ 2012; 344:e1552. (41) Smith GC. Researching new methods of screening for adverse pregnancy outcome: lessons from pre-eclampsia. PLoS Med 2012; 9(7):e1001274.
*Studies of particular importance to the current topic.
ACCEPTED MANUSCRIPT Table 1. Top 11 research priorities from a stillbirth priority setting partnership
TE D
M AN U
SC
RI PT
Research questions How can the structure and function of the placenta be assessed during pregnancy to detect potential problems and reduce the risk of stillbirth? Does ultrasound assessment of fetal growth in the third trimester reduce stillbirth? Do modifiable ‘lifestyle’ factors (e.g. diet, vitamin deficiency, sleep position, sleep apnea, lifting and bending) cause or contribute to stillbirth risk? Which investigations identify a fetus at risk of stillbirth after a mother believes she has experienced reduced fetal movements? Can the wider use of existing tests and monitoring procedures, especially in later pregnancy, and the development and implementation of novel tests (biomarkers) in the mother or in early pregnancy, help prevent stillbirth? What causes stillbirth in normally grown babies? What is the most appropriate bereavement and postnatal care for both parents following a stillbirth? Which antenatal care interventions are associated with a reduction in the number of stillbirths? Would more accessible evidence-based information on signs and symptoms of stillbirth risk, designed to empower women to raise concerns with healthcare professionals, reduce the incidence of stillbirth? How can staff support women and their partners in subsequent pregnancies, using a holistic approach to reduce anxiety, stress and any associated increased visits to healthcare settings? Why is the incidence of stillbirth in the UK higher than in other similar high-income countries, and what lessons can we learn from this?
AC C
EP
Text quoted verbatim from Heazell et al, 2015 (reference 3)
ACCEPTED MANUSCRIPT Table 1. Maternal risk factors for stillbirth. 95% Adjusted Odds ratio confidence interval Non-Hispanic black race/ethnicity 2.12 1.41 to 3.20 Previous stillbirth 5.91 3.18 to 11.00 Nulliparity + previous losses at <20 weeks' 3.13 2.06 to 4.75 Nulliparity, no previous losses 1.98 1.51 to 2.60 Diabetes mellitus 2.50 1.39 to 4.48 Maternal age 40 years or older 2.41 1.24 to 4.70 Maternal AB blood type 1.96 1.16 to 3.30 History of drug addiction 2.08 1.12 to 3.88 Smoking 1.55 1.02 to 2.35 Obesity/overweight 1.72 1.22 to 2.43 Not living with a partner 1.62 1.15 to 2.27 Multiple pregnancy 4.59 2.63 to 8.00
M AN U
SC
RI PT
Characteristic
AC C
EP
TE D
Data from reference 5 (see publication for details of multivariate analysis and referent categories). Stillbirth defined on basis of ≥20 weeks gestational age threshold.
ACCEPTED MANUSCRIPT Table 2. Maternal risk factors for gestational diabetes mellitus
AC C
EP
TE D
M AN U
SC
RI PT
Maternal risk factors for gestational diabetes mellitus Body mass index >30 kg/m2 Previous macrosomic infant (>4.5kg) Previous gestational diabetes First degree relative with diabetes Ethnic origin associated with high incidence of diabetes
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
ACCEPTED MANUSCRIPT Highlights
•
Stillbirth in the single major determinant of perinatal death
•
Screening for stillbirth currently targets investigations to women with risk factors, which are poorly discriminative There is scope for screening using ultrasound and biomarkers
•
The primary disease modifying intervention is delivery
•
Induction at term prevents perinatal death without increasing caesarean risk
AC C
EP
TE D
M AN U
SC
RI PT
•
PII:
S1521-6934(16)30075-X
DOI:
10.1016/j.bpobgyn.2016.08.002
Reference:
YBEOG 1634
To appear in:
Best Practice & Research Clinical Obstetrics & Gynaecology
Received Date: 4 May 2016 Revised Date:
26 August 2016
Accepted Date: 30 August 2016
Please cite this article as: Smith GCS, Screening and prevention of stillbirth, Best Practice & Research Clinical Obstetrics & Gynaecology (2016), doi: 10.1016/j.bpobgyn.2016.08.002. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
1
ACCEPTED MANUSCRIPT Screening and prevention of stillbirth
Prof Gordon C.S. Smith, DSc, FMedSci. Professor & Head of Department, Obstetrics and Gynaecology, University of Cambridge; NIHR
RI PT
Cambridge Comprehensive Biomedical Research Centre, Cambridge, CB2 0SW, UK.
Correspondence to:
Prof GCS Smith, Department of Obstetrics and Gynaecology, University of Cambridge, Box 223 The Rosie Hospital, Cambridge, CB2 2SW, UK.
AC C
EP
TE D
M AN U
SC
Tel: +44 (0)1223 336871, Fax: +44 (0)1223 215327, E-mail: [email protected]
2
ACCEPTED MANUSCRIPT Abstract (150 words)
Stillbirth is delivery of a baby at or after 24 weeks of gestational age (UK definition) showing no signs of life. It affects almost 1 in 200 pregnancies and is the single major cause of perinatal death. Stillbirth is associated with a wide range of maternal demographic characteristics, but most of the
RI PT
variation in stillbirth risk is independent of these characteristics. Stillbirth is the end point of multiple processes, but the single most common cause is probably placental dysfunction. Stillbirth is
associated with a range of biochemical and ultrasonic predictors, but there is limited evidence to support population based screening. However, the evidence based is weak due by use of poorly characterised screening tests, the failure to couple risk assessment with a clearly effective
SC
intervention for those who screen positive, and inadequate study sample sizes. Basic research needs
M AN U
to identify better predictors, and clinical trials need to adopt more rigorous methodology.
Key words:
AC C
EP
TE D
Stillbirth, screening, intervention, ultrasound, biomarkers.
3
ACCEPTED MANUSCRIPT 1. Introduction The 2016 Lancet Ending Preventable Stillbirth Series indicated that the UK was in about the middle of the range of 49 high income countries in relation to stillbirth rates, and had one of the slowest rates of decline in stillbirth.1 The absolute risk of stillbirth from 24 weeks onwards is 3-5 per 1,000, i.e. about the same as the total risk of death in the first year of life. Moreover, a significant proportion of
RI PT
stillbirths potentially could have been prevented if the babies had been identified as high risk, i.e. cases where the cause of death was not a major congenital anomaly and where the death occurred at a gestational age associated with a low risk of infant mortality.2 A recent study, supported by the James Lind Alliance, reported the top research priorities in relation to stillbirth. This study elicited 1275 responses from 574 participants (equally divided between professionals and non-professionals)
SC
and identified several priorities (out of 300 indicative unanswered questions) directly related to screening, (e.g. Priority #5, Does ultrasound assessment of fetal growth in the third trimester reduce
M AN U
stillbirth?), and 4 others which were also relevant (e.g. Priority #6, Would increasing the frequency of umbilical artery Doppler scanning during pregnancy reduce stillbirth?).3 The top 11 recommendations are listed (Table 1). The aim of the current review is to consider current practice in relation to screening for stillbirth, and how improved methods of screening may be developed and evaluated.
TE D
1.1 Epidemiology of stillbirth
Stillbirth is defined as delivery of a baby in the perinatal period which fails to show any signs of life. The definition of the start of the perinatal period varies between countries. In the USA it is 20 weeks of gestational age (wkGA) and in the UK it is 24wkGA.2 The rate of stillbirth in high income countries
EP
varies from 1.3 per 1,000 to 8.8 per 1,000, with an average of 3.5 per 1,000.1 The clear majority of stillbirths in high income countries are the result of intra-uterine fetal death prior to the onset of
AC C
labour (antepartum stillbirth).2 Deaths occurring during labour (intrapartum stillbirths) account for 510% of all stillbirths in high income countries but account for a much larger proportion of losses in low and middle income countries (LMIC).4 A range of maternal characteristics have been associated with the risk of stillbirth. One of the most systematic and methodologically sound analyses of such risk factors was performed by the NICHD's Stillbirth Collaborative Research network.5 Their analysis of risk factors identifiable at the start of pregnancy is tabulated (Table 2).
1.2 Causes and classification of stillbirth Aside from the sub-division of stillbirths into intrapartum and antepartum, losses can also be classified according to the presumed cause. However, this process is complicated by the fact that
4
ACCEPTED MANUSCRIPT relatively few losses have a completely understood cause of death. The remainder have varying degrees of uncertainty in the mechanism(s) leading to death. This is illustrated in Figure 1 in relation to ascribing cause of death in the presence of a range of maternal medical conditions. A diverse – if not bewildering – array of classification systems have been developed.1 One of the main characteristics which determines variation between the systems is the extent to which they
RI PT
will attribute a given associated condition or finding as being the cause of death. For example, an unexpected and unexplained stillbirth of an infant at 39 weeks where the baby's birth weight was on the 2rd percentile for sex and gestational age might be defined as unexplained in one classification system and as being due to fetal growth restriction (FGR) in another. In a sense, both classifications are correct. The actual cause of death is unknown, hence the loss is strictly unexplained. However, it
SC
is known that birth weight <3rd percentile is associated with a 10-fold risk of stillbirth at term.6
Hence, it is very likely that the baby's death was related to its size. The subject of classification of
M AN U
stillbirth is reviewed in detail elsewhere.7
In the context of screening, one of the key associations for stillbirth is poor fetal growth. It is estimated that between 30 and 50% of stillbirths are associated with low birth weight percentile, and this is assumed to reflect FGR, which is in turn presumed to be related, in a large proportion of cases, to placental dysfunction.2 This association is important in the context of screening, as a
2. Screening for stillbirth
TE D
number of tools exist to quantify both fetal growth and placental function.
The first element of screening is to differentiate a population into those at high risk of the condition and those at low risk. The analysis and interpretation of screening statistics can be complicated.
EP
Particular problems in the context of stillbirth are as follows. (i) stillbirth is the end result of diverse pathological processes. Hence, no one test is likely to be
AC C
highly sensitive for the condition overall. (ii) stillbirth can occur across the whole range of gestational age. As the primary means to prevent stillbirth is to deliver the baby, the consequences of inappropriate intervention differ profoundly: at 24 weeks outcomes are generally extremely poor, whereas at 40 weeks, outcomes are generally extremely good.
(iii) stillbirth is relatively uncommon, hence, any sub-type of stillbirth is less common still. For example, if 40% of stillbirths are due to FGR and the risk of stillbirth in a country is 4 per 1,000, the absolute risk of stillbirth associated with FGR is 1.6 per 1,000. Normally, a test with a positive likelihood ratio of 20 would be regarded as an excellent screening test. However, the positive predictive value of this test in such a population would be about 3%. Consequently, 97% of screen
5
ACCEPTED MANUSCRIPT positive women would not experience the event of interest. It follows, therefore, that if an intervention based on such a test caused any harm to false positives, that a programme of screening and intervention would be likely to cause harm despite a strong screening test and an effective intervention.
RI PT
3. Maternal risk factors The maternal characteristics which are associated with the risk of stillbirth have been tabulated (Table 2). However, collectively, these associations explained just 19% of the variability in the risk of stillbirth. Hence, programmes of screening and intervention which focus on maternal risk factors will have a relatively limited capacity to reduce the numbers of stillbirth. Another key element of
SC
maternal risk are the associations with gestational diabetes mellitus (GDM). There are a series of maternal characteristics which place a woman at increased risk of GDM (Table 3). A recent detailed
M AN U
review of a representative sample of unexpected stillbirths of normally formed infants at term in the UK, conducted by the national perinatal mortality surveillance system, MBRRACE-UK, found that failure to screen for GDM was one of the three most common potentially preventable characteristics observed which might have resulted in the loss being avoided.8
3.1 Clinical factors associated with stillbirth risk
TE D
Reduced fetal movements (RFM) is one of the key symptoms associated with the risk of stillbirth. This subject has been reviewed in detail elsewhere.9 Failure of medical and midwifery staff to respond appropriately to RFM was one of the other major potentially preventable causes of stillbirth in the MBRRACE-UK review described above.8 The RCOG in the UK has issued a guideline for the
EP
management of women who present with reduced movements.10 Another major symptom associated with stillbirth risk is antepartum haemorrhage. In the absence of placenta praevia or
AC C
cervical pathology, this is likely to represent bleeding from the placenta, and may be a harbinger of later abruption. Finally, acquired disorders of pregnancy (such as GDM or pre-eclampsia) could be associated with maternal symptoms. Management is based on specific testing for the condition in question.
3.2 Response to women with symptoms associated with stillbirth risk At preterm gestational ages, fetal monitoring is indicated. Generally, the first line of assessment is cardiotocography (CTG, also called a non-stress test in the USA). Interestingly, the RCOG Guideline on RFM does not recommend computerised CTG,10 which utilises objective analysis of the trace, including computerised assessment of beat to beat variability. However, the Cochrane reviews
6
ACCEPTED MANUSCRIPT indicate no benefit of using non-computerised CTG (and a trend towards harm) compared with using nothing, and a reduced risk of perinatal death using computerised CTG analysis compared with noncomputerised CTG (see Smith 2015, for review).11 Hence, computerised CTG is a reasonable first step. The next level of assessment is an ultrasound scan. This is done if there are other risk factors present, but will also be done in low risk women who are presenting with their second (or greater)
RI PT
episode of RFM. The key evidence-based measurement in high risk pregnancies is Doppler flow velocimetry of the umbilical artery. However, generally, fetal biometry and amniotic fluid
measurement would also be performed, sometimes with additional Doppler measurements (middle cerebral artery). The evidence base supporting these supplementary measurements is poor: this is a key area for further research and multiple recommendations of the prioritisation exercise for
SC
stillbirth research touched on this area.
M AN U
At term, induction of labour should be considered for women presenting with RFM. At 37-38 weeks, induction is still associated with increased risks of perinatal morbidity and should only be offered if fetal assessment is non-reassuring, there are other risk factors or if there have been multiple episodes of RFM. However, perinatal outcome is optimal at 39-41 weeks and induction should be considered for any women presenting with RFM at ≥39 weeks. Epidemiological evidence exists that universal induction at 39 weeks would reduce overall rates of stillbirth.12 However, for women with a
TE D
single episode of RFM, who have no other risk factors, and who want to avoid intervention, it is reasonable not to induce labour, as the absolute risk of stillbirth is likely to be <1%. The basis for this statement is that the background risk of stillbirth at term is 1-2 per 1,000 and it is unlikely that a
EP
single episode of RFM in an otherwise low risk women is associated with a >5-fold risk of stillbirth.
4. Biochemical predictors of stillbirth risk
AC C
A range of biochemical tests have been associated with the risk of stillbirth. Most of these are maternal blood tests but, historically, urine tests were also evaluated. Many of the associations described were secondary analyses of measurements made for the purposes of screening for unrelated conditions, such as neural tube defects and Down's syndrome.
4.1 Pregnancy associated plasma protein A (PAPP-A) PAPP-A is a protease for insulin like growth factor binding proteins 4 and 5, and it is produced by the placenta. Its primary application has been for screening for Down's syndrome. However, it was shown that low first trimester PAPP-A was associated with stillbirth risk and that the association was due to losses related to placental dysfunction.13;14 A systematic review concluded that PAPP-A was a
7
ACCEPTED MANUSCRIPT "good predictor of stillbirth related to placental dysfunction disorders".15 In the UK, an RCOG Guideline recommends methods for further monitoring of women found to have low PAPP-A in the first trimester of pregnancy.16
4.2 Alpha fetoprotein (AFP)
RI PT
AFP is a fetal oncotic protein. High maternal serum levels are associated with structural fetal anomalies (neural tube defects and abdominal wall defects) and low levels are associated with
Down's syndrome. High levels in the second trimester are associated with the risk of stillbirth in normally-formed babies, but the associations are stronger for losses occurring at extreme preterm gestational ages.17 Hence, intervention to prevent such losses is problematic. The association is
SC
thought to be related to abnormal permeability of the placenta, hence increased passage of the protein across the placental barrier. Interestingly, unexplained raised maternal serum AFP is
where the pregnancy was complicated.18
4.3 Others
M AN U
associated with a diverse series of abnormalities on subsequent placental pathological examination
A range of other biochemical tests have been associated with the risk of FGR and stillbirth. Some of these were in widespread use prior to the advent of ultrasound, such as measurement of oestrogens
TE D
in maternal urine. Others have been more recently identified as potential screening tests for preeclampsia, in particular pro- and anti-angiongenic proteins circulating in the mother’s blood, such as PlGF, sFlt-1, and eEng. Studies have described some promising associations with the angiogenic regulators.19 However, the relative rarity of stillbirth (compared with preeclampsia and preterm
EP
birth) make it a more difficult subject to address. Cohort studies have to be larger in order to be able to detect associations. Moreover, if a measurement has a positive likelihood ratio of 10, which
AC C
would usually be regarded as a good screening test, it would only have a positive predictive value of <5%, and much lower still if performed later in pregnancy.
Biochemical tests have great potential as a screening tool for stillbirth when combined with ultrasound. However, any such potential will not be adequately evaluated unless studies are sufficiently powered to detect associations. Moreover, interventional studies will present huge challenges and will require addressing significant methodological considerations (see below).
5. Ultrasonic assessment of stillbirth risk
8
ACCEPTED MANUSCRIPT The first use of ultrasound to estimate the weight of the fetus was described in 1975.20 Subsequently, multiple other methods were described to assess fetal well-being, including biophysical measurements (activity, breathing movements, tone and liquor volume) and ultrasonic Doppler flow velocimetry of fetal and utero-placental arteries and veins. Research focused on both the diagnosis of fetal compromise in women who had presented with complications and/or those
RI PT
known to have risk factors, and also on trying to identify which apparently low risk women had occult fetal compromise. The body of work quickly led to RCTs of universal screening of low risk women using late pregnancy ultrasound. A meta-analysis of these trials (latest version is Bricker et al 201521 including data from a total of 34,980 women recruited to 13 trials), reported no maternal or fetal benefit of screening. A previous version of this meta-analysis (which had very similar
M AN U
not have routine late pregnancy ultrasound scans.
SC
conclusions) led NICE to conclude in their 2008 Guideline on Antenatal Care22 that women should
In the same Guideline, NICE indicated that further studies on the diagnostic effectiveness of universal ultrasound as a screening method in the general population should be considered a priority. It may seem counterintuitive that NICE would declare universal screening to be ineffective but then to propose further studies on the diagnostic effectiveness of the screening test. This
TE D
apparent anomaly is explained by closer examination of the studies included in the meta-analysis. Ultimately, all meta-analyses, however appropriately conducted, are only as good as the trials included. Issues with the 13 trials in the Cochrane review are as follows: (i) the individual trials used different definitions of screen positive, (ii) none of the definitions had been evaluated in a level 1
EP
study of diagnostic effectiveness prior to being employed in the screening study, i.e. all of the individual interventional studies were designed and conducted in the absence of high quality
AC C
quantitative information about the performance of the screening test, (iii) none of the trials had a standardised intervention for those who screened positive, other than to reveal the test or to perform further scans. Finally, even the meta-analysis itself is under-powered. The 2016 Lancet Ending Preventable Stillbirth Series reported sample size calculations for trials of screening and intervention for stillbirth. These looked at the statistical power assuming very optimistic values of screening test performance (5% of the population identified with a positive likelihood ratio of 10) and intervention (50% reduction) based on the incidence of stillbirth (1 in 200). The sample size required (90% power, alpha 0.05, two sided) was 131,000.1 Hence, the meta-analysis has only about a quarter of the sample size required for even a very good screening test coupled with a highly effective intervention. Consequently, despite the fact that ultrasound has been used for >40 years to assess fetal well-being, and despite the fact that it is the primary method for assessing and
9
ACCEPTED MANUSCRIPT monitoring high risk pregnancies, there is an absence of evidence in relation to its utility in screening low risk women. Universal ultrasound has been assessed as a means of detecting small for gestational age (SGA) infants in a level 1 study of diagnostic effectiveness (i.e. where the result of the research scans was blinded) including ~4,000 first pregnancies.23 The rate of detection of SGA infants was 20% with
RI PT
selective use of ultrasound and 57% with universal scanning. However, for every one additional true positive result, there were two additional false positives with universal scanning. Moreover, of a range of ultrasonic indicators of FGR, only one differentiated between SGA infants and the risk of perinatal morbidity. Fetuses which were SGA on scan and had reduced growth velocity of the
SC
abdominal circumference (compared with the 20 week scan) were at increased risk of morbidity compared with normal sized fetuses, whereas SGA was not associated with the risk of morbidity if the growth velocity was normal. A meta-analysis has also demonstrated that high resistance
M AN U
patterns of Doppler flow velocimetry in the uterine arteries in mid-gestation are a good predictor of stillbirth related to placental dysfunction.15 However, the studies revealed the result of the scan, which complicates assessment of the association.
Many private ultrasound providers have started offering late pregnancy ultrasound scans. There is, therefore, ad hoc screening at present with very little evidence regarding the balance of risks and
TE D
benefits. Late pregnancy ultrasound has the potential to cause harm. False positive diagnoses can lead to unnecessary intervention and, in particular, early term delivery. While this is a valuable and evidence based intervention in certain contexts, such as pre-eclampsia, infants born at 37-38 weeks are at increased risk of short term morbidity,24 and even have poorer achievement at school.25
EP
Hence, ad hoc screening and intervention has the potential to cause short and long term harm through iatrogenic late preterm and early term birth. This has recently been recognised in France,
AC C
where routine ultrasound was implemented in the absence of supportive evidence. Screening did not appear to confer any benefit, but false positive diagnosis of SGA was associated with a greater than 3-fold increase in the risk of medically-indicated preterm birth.26 This does not indicate that universal screening is futile. However, these findings underline the importance of understanding what elements of prenatal ultrasonic assessment of the fetus in a low risk pregnancy are informative of risk, and what interventions can be applied to mitigate the risk without causing harm.
6. Interventions to prevent stillbirth 6.1 Medical interventions 6.1.1 Aspirin
10
ACCEPTED MANUSCRIPT Low dose aspirin has been widely evaluated as a method for preventing placentally-related complications in pregnancy and, in particular, pre-eclampsia. Aspirin acts by irreversibly inhibiting the enzyme cyclo-oxygenase (COX), which is a key point in the biochemical synthesis of prostanoids, including thromboxane. Aspirin selectively blocks platelet prostanoid production as they lack a nucleus and cannot synthesise further COX following the irreversible binding with aspirin. The Cochrane review contains 14 trials containing 33,098 women, and demonstrates an 14% reduction in
RI PT
the risk of stillbirth or neonatal death.27 Hence, aspirin should be considered for all women at risk of stillbirth. Typically, it would be started at 12 weeks and discontinued at 36 weeks. There is
preliminary evidence from the pre-eclampsia literature to suggest that starting the treatment prior to 16 weeks may be associated with greater risk reduction.28
SC
6.1.2 Low molecular weight heparin
Given the success of aspirin and given some published associations between maternal thrombophilia
M AN U
and stillbirth risk, many groups have considered the use of LMWH as a preventative treatment for women at increased risk of stillbirth. It is regarded as a strongly indicated therapy in some situations, such as anti-phospholipid syndrome, where it has the additional benefit of reducing the risk of maternal venous thrombo-embolism. A series of small trials appeared to show a benefit of the treatment on perinatal outcome.29 However, a subsequent multicentre study demonstrated no beneficial effect of LMWH.30 A caveat to this is that even that study was only powered to detect very
TE D
major effects of LMWH. The confidence intervals around the effect from the trial include reductions in stillbirth risk which would be regarded as clinically significant. The current status of LMWH is, therefore, that it’s use should primarily be considered to prevent maternal venous thromboembolism and that further studies are required before its use is recommended for the primary aim
EP
of preventing fetal complications.
AC C
6.1.3 Nitric oxide donors and selective phosphodiesterase inhibitors Nitric oxide, a gaseous mediator of smooth muscle relaxation, is thought to have a key role in the control of placental development, with a deficit of nitric oxide effects postulated to be a determinant of abnormal placentation.31 This has led to studies evaluating methods for increasing nitric oxide. Broadly, these fall into two categories: (i) nitric oxide donors, and (ii) selective phosphodiesterase inhibitors. Many nitric oxide donors have been developed for treatment of cardiovascular disease, such as angina and heart failure. A number of studies have evaluated nitric oxide donors in high risk populations. These have generally been small scale single centre studies and some have generated positive results.32;33 However, further large scale multicentre trials are required prior to introduction of these agents in the management of high risk women. Selective
11
ACCEPTED MANUSCRIPT phosphodiesterase inhibitors were first developed for cardiovascular applications. But early experience in healthy volunteers led to an unanticipated effect on male erectile function, and a range of these drugs is available for erectile dysfunction. The best known is sildenafil citrate (known commercially as Viagara), and the evidence supporting a beneficial effect of this drug in the context of early onset FGR has been reviewed.34 Currently, a multicentre RCT is in process, STRIDER, which
RI PT
seeks to assess the effect of sildenafil citrate in pregnancies complicated by early onset severe FGR. The field may also develop into the evaluation of other selective PDE inhibitors with longer half-life, and the use of the drug in pregnancies at high risk of FGR by their screening results, but prior to the disease onset.
SC
6.1.4 Others
A range of other specific therapies exist, which are at various stages of evaluation but are not likely
M AN U
to be in routine clinical use in the very near future. These include the use of supplemental oxygen35 and gene therapy.36 Both of these approaches face significant logistical issues. Gene therapy has been evaluated in animal models and is currently being assessed for acceptability and feasibility in the context of severe, early onset FGR.
Finally, optimal management of co-existing medical conditions is an indirect way to reduce the risk
TE D
of stillbirth associated with maternal disease. The obvious example is diabetes mellitus, both gestational and pre-existing. The risk of perinatal mortality reduces with better control. Although diabetes is the obvious example, it is plausible that optimal medical therapy would improve perinatal
disorders etc.
EP
outcome with other maternal medical conditions, such as thyroid disease, connective tissue
AC C
6.2 Delivery to prevent stillbirth
At present, the only intervention which has a major effect on the risk of stillbirth is delivery. The effect of this clearly depends on (i) the background risk of stillbirth, and (ii) the gestational age, as GA is the major determinant of the risk of neonatal death. At term, the risk of neonatal death is extremely low, particularly following exclusion of deaths due to anomalies. However, the weekly risk of stillbirth is stable from 24 onwards and then rises at term and postterm. Combining analysis of declining risk of neonatal death and increasing and cumulative risk of stillbirth, it has been estimated that the risk of perinatal death is lowest at around 39 weeks.12 Hence, one approach to preventing term stillbirths would simply be to induce labour in all women at 39 weeks. This approach is supported by a Cochrane review, which indicates that routine induction of labour at term and post
12
ACCEPTED MANUSCRIPT term reduces the risk of perinatal mortality by ~70%.37 However, most would feel that this level of intervention would be excessive in relation to the number of deaths prevented. Nevertheless, this background should inform attitudes in general. If a woman feels strongly that she should be induced at 39 weeks or later, it is probably not unreasonable to agree to the request assuming the healthcare system is sufficiently well resourced. Furthermore, a low threshold for induction of labour should be
RI PT
applied to women with concerning symptoms or investigations. If a woman presents with reduced fetal movements (RFM) at 34 weeks, induction of labour would only be considered if there were confirmatory tests indicating other concerns. However, at 39 weeks and beyond, it would be entirely reasonable to induce labour for RFM. This would be especially the case if the woman actively wanted induced, and it would be appropriate if the woman was neutral and expressed the sentiment of so
SC
many “I just want what is best for my baby”. However, many women are keen to avoid induction. In such cases, an open discussion is required. Women who decline induction can be reassured that,
M AN U
with a single episode of RFM, no associated risk factors, and reassuring monitoring, the absolute risk of a stillbirth remains low. It is essential in all these discussions that the woman’s own attitudes and preferences are a foundation for the discussion. A key element for any such discussion would be to contradict the misconception that induction of labour increases the risk of caesarean section. A systematic review indicates that it results in a lower risk of caesarean delivery.38 Moreover, a recent RCT has shown no increase in the risk of caesarean section when labour was induced at 39 weeks in
greater).39
TE D
a group of women with a high background risk of caesarean (nulliparous women aged 35y or
6.3 Modifying maternal risk factors for stillbirth
EP
A number of the maternal characteristics associated with stillbirth risk are potentially modifiable. Numerous studies have demonstrated higher rates of stillbirth among smokers, and a causal
AC C
association is plausible. It follows, therefore, that smoking cessation is likely to reduce the risk of stillbirth. Given the problems of conducting RCTs of interventions to prevent stillbirth (see below), it is unlikely that there will ever be level 1 evidence to support the use of smoking cessation as a means to prevent stillbirth. Nevertheless, given the background and given that this represents an absence of evidence, rather than evidence of absence, recommending that mothers participate in smoking cessation would be an appropriate approach. Similarly, raised body mass index is both directly associated with stillbirth in epidemiological studies and is also a risk factor for a cause of stillbirth, namely, GDM. Weight reduction should be considered for women who are obese and are planning pregnancy. Dieting through calorific restriction is controversial in pregnancy. However, women who are obese should be referred to a dietician and approaches which might improve
13
ACCEPTED MANUSCRIPT outcomes include reducing quantities of sugar intake, controlling portion sizes, eating more fresh fruit and vegetables, and the maintenance of healthy levels of activity through pregnancy. A number of epidemiological studies have shown associations between maternal sleep position and the risk of stillbirth. These have demonstrated increased risks of stillbirth with supine sleep position, and it is speculated that this may be due to reduced blood supply to the uterus through caval
RI PT
compression (i.e. a pressure effect of the gravid uterus on the inferior vena cava). Further studies aiming to confirm or refute the association are in progress. If these confirm the association with maternal sleep position, there may be a case for a public health campaign to modify maternal sleep position as a means to reduce stillbirth risk. This is analogous with sudden infant death syndrome: epidemiological studies demonstrated an association with prone infant sleeping positon and SIDS,
SC
and a public health campaign was followed by an extraordinarily rapid decline in rates of SIDS in the
M AN U
UK.40
7. Evaluating novel methods for screening and prevention of stillbirth. Despite huge expenditure of time and resources, there is no clear approach to screening women for stillbirth risk. Numerous studies have described possible methods for screening, including biochemical tests and ultrasound scans. However, none has been clearly shown to be effective in randomised controlled trials. The negative results of the trials do not necessarily indicate that
TE D
ultrasound is futile. As discussed above, there are multiple issues in relation to the meta-analysis of RCTs of universal ultrasound. It is essential, therefore, that any future trials learn the lessons from previous trials and address new methods of screening using a methodologically rigorous approach. Key elements in designing new RCTs to prevent stillbirth are as follows: (i) the eligible population, (ii)
EP
the screening test to be employed and its diagnostic effectiveness, (iii) the timing of randomisation, namely, before or after application of the screening test, (iv) the intervention, (v) the control group,
AC C
(vi) the primary and secondary outcomes, (vii) the sample size, (viii) the possibility of randomising at the level of hospitals, e.g. using a cluster RCT or stepped wedge RCT, (ix) the potential use of an adaptive trial design, (x) whether an initial smaller scale feasibility study is required, and (xi) the feasibility of the study, including manpower, skills mix, and quality assurance of the screening test. Many of these issues have previously been discussed in more depth elsewhere.41
8. Summary Stillbirth in the single major determinant of perinatal death. In high income countries, ≥90% follow death of the baby prior to the onset of labour, and in this setting typical absolute rates are 3-4 per 1,000. Stillbirth is the end result of diverse processes, hence a single perfect screening test is unlikely
14
ACCEPTED MANUSCRIPT to be developed. Currently, women are assessed for their risk of stillbirth at the time of first attendance for antenatal care, and then are re-assessed during the pregnancy either at routine visits or when they present with symptoms which are associated with increased risk of stillbirth (such as reduced fetal movements). A range of methods have been developed to estimate the risk of stillbirth and associated conditions (such as FGR). Currently, investigations such as ultrasound are only
RI PT
applied to women who have risk factors or attend with complications. However, the possible utility of universal ultrasound has been poorly evaluated. About one third of all stillbirths occur at term and these are, potentially, the most readily preventable as there is a safe and effective intervention, namely, induction of labour, which reduces the risk of perinatal death by 70%. A key challenge is how to identify the women at increased risk of stillbirth and who would benefit from this
M AN U
causing unacceptably high rates of intervention.
SC
intervention, from the general healthy population. In this way, losses could be prevented without
Conflicts of interest.
GS receives/has received research support from GE, Roche and GSK. GS has been paid to attend advisory boards by GSK and Roche. GS has acted as a paid consultant to GSK. GS has received support to attend a scientific meeting from Chiesi. GS is named inventor in a patent submitted by
TE D
GSK (UK), for novel application of an existing GSK compound for the prevention of preterm birth (PCT/EP2014/062602). GS has acted as an expert witness. GS is a member of a Data Safety
Practice points •
EP
Monitoring Committee for a trial of an RSV vaccine in pregnancy, being run by GSK.
Maternal risk factors do not perform well as a screening test for stillbirth. Many stillbirths
•
AC C
occur to low risk women
Common issues with quality of care in relation to stillbirth prevention include failure of
effective screening or treatment of GDM, reduced fetal movements, or poor fetal growth
•
Non-computerised antenatal cardiotocography has little role in assessing fetal well being,
and may provide false reassurance about stillbirth risk in women presenting with reduced movements •
Induction of labour can safely be considered from 39 weeks onwards, even in relatively low risk women. It will prevent subsequent stillbirth, and is not associated with an increased risk of caesarean delivery.
15
ACCEPTED MANUSCRIPT •
Women at risk of stillbirth should be treated with low dose aspirin from 12 to 36 weeks. Low molecular weight heparin should primarily be used to prevent maternal thrombo-embolism and its use to prevent stillbirth will largely be confined to randomised controlled trials
• Research agenda Design, feasibility and optimal design of an RCT of universal ultrasound, and the effect of
RI PT
•
universal ultrasound scanning •
The best ultrasonic markers of likely intra-uterine fetal death in women presenting with complications
•
The ability of biomarkers to differentiate between women with normal and abnormal
•
SC
placental function
The relationship between placental pathology and adverse pregnancy outcome, and
M AN U
methods for diagnosing specific placental pathological processes in the antenatal period (e.g. blood tests for recognised placental disease processes)
EP
TE D
The mechanisms leading to placental dysfunction
AC C
•
16
ACCEPTED MANUSCRIPT Legend for Figure.
AC C
EP
TE D
M AN U
SC
RI PT
Figure 1. Representation of the continuum of uncertainty in ascribing cause for stillbirth, using the example of losses associated with maternal medical conditions. ALT denotes alanine transaminase; GA denotes gestational age. Reproduced with permission from a report of an NICHD workshop (Reddy et al., Obstet Gynecol 2009;114:901-914).
17
ACCEPTED MANUSCRIPT References
*(1) Flenady V, Wojcieszek AM, Middleton P, Ellwood D, Erwich JJ, Coory M et al. Stillbirths: recall to action in high-income countries. Lancet 2016; 387(10019):691-702. *(2) Smith GC, Fretts RC. Stillbirth. Lancet 2007; 370(9600):1715-1725.
RI PT
(3) Heazell AE, Whitworth MK, Whitcombe J, Glover SW, Bevan C, Brewin J et al. Research priorities for stillbirth: process overview and results from UK Stillbirth Priority Setting Partnership. Ultrasound Obstet Gynecol 2015; 46(6):641-647.
SC
(4) Lawn JE, Blencowe H, Pattinson R, Cousens S, Kumar R, Ibiebele I et al. Stillbirths: Where? When? Why? How to make the data count? Lancet 2011; 377(9775):1448-1463. *(5) SCRN Writing Group . Association between stillbirth and risk factors known at pregnancy confirmation. JAMA 2011; 306(22):2469-2479.
M AN U
(6) Moraitis AA, Wood AM, Fleming M, Smith GC. Birth weight percentile and the risk of term perinatal death. Obstet Gynecol 2014; 124(2 Pt 1):274-283. (7) Flenady V, Froen JF, Pinar H, Torabi R, Saastad E, Guyon G et al. An evaluation of classification systems for stillbirth. BMC Pregnancy Childbirth 2009; 9:24.
TE D
(8) Draper ES, Kurinczuk JJ, Kenyon S. MBRRACE-UK Perinatal confidential enquiry: term, singleton, normally formed, antepartum stillbirth. Leicester, UK: The Infant Mortality and Morbidity Studies, Department of Health Sciences, University of Leicester; 2015. (9) Froen JF, Tveit JV, Saastad E, Bordahl PE, Stray-Pedersen B, Heazell AE et al. Management of decreased fetal movements. Semin Perinatol 2008; 32(4):307-311.
EP
(10) RCOG. RCOG Green-top guideline No. 57: Reduced fetal movements. London, UK: RCOG; 2011. (11) Smith GCS. Prevention of stillbirth. The Obstetrician and Gynaecologist 2015;DOI 10.1111/tog.12197.
AC C
(12) Smith GC. Life-table analysis of the risk of perinatal death at term and post term in singleton pregnancies. Am J Obstet Gynecol 2001; 184(3):489-496. (13) Smith GC, Stenhouse EJ, Crossley JA, Aitken DA, Cameron AD, Connor JM. Early pregnancy levels of pregnancy-associated plasma protein a and the risk of intrauterine growth restriction, premature birth, preeclampsia, and stillbirth. J Clin Endocrinol Metab 2002; 87(4):1762-1767. (14) Smith GC, Crossley JA, Aitken DA, Pell JP, Cameron AD, Connor JM et al. First-trimester placentation and the risk of antepartum stillbirth. JAMA 2004; 292(18):2249-2254. *(15) Conde-Agudelo A, Bird S, Kennedy S, Villar J, Papageorghiou A. First- and second-trimester tests to predict stillbirth in unselected pregnant women: a systematic review and metaanalysis. BJOG 2014.
18
ACCEPTED MANUSCRIPT (16) RCOG. Guideline No. 31: The investigation and management of the small for gestational age fetus. 1-34. 2013. London, UK, RCOG Press.
(17) Smith GC, Shah I, White IR, Pell JP, Crossley JA, Dobbie R. Maternal and biochemical predictors of antepartum stillbirth among nulliparous women in relation to gestational age of fetal death. BJOG 2007; 114(6):705-714.
RI PT
(18) Jauniaux E, Moscoso G, Campbell S, Gibb D, Driver M, Nicolaides KH. Correlation of ultrasound and pathologic findings of placental anomalies in pregnancies with elevated maternal serum alpha-fetoprotein. Eur J Obstet Gynecol Reprod Biol 1990; 37(3):219-230.
SC
(19) Chaiworapongsa T, Romero R, Korzeniewski SJ, Kusanovic JP, Soto E, Lam J et al. Maternal plasma concentrations of angiogenic/antiangiogenic factors in the third trimester of pregnancy to identify the patient at risk for stillbirth at or near term and severe late preeclampsia. Am J Obstet Gynecol 2013; 208(4):287.
M AN U
(20) Campbell S, Wilkin D. Ultrasonic measurement of fetal abdomen circumference in the estimation of fetal weight. Br J Obstet Gynaecol 1975; 82(9):689-697. *(21) Bricker L, Medley N, Pratt JJ. Routine ultrasound in late pregnancy (after 24 weeks' gestation). Cochrane Database Syst Rev 2015; 6:CD001451. (22) National Collaborating Centre for Women's and Children's Health. NICE Guideline: Antenatal care. 2008. London, RCOG Press.
TE D
*(23) Sovio U, White IR, Dacey A, Pasupathy D, Smith GCS. Screening for fetal growth restriction with universal third trimester ultrasonography in nulliparous women in the Pregnancy Outcome Prediction (POP) study: a prospective cohort study. Lancet 2015; 386:2089-2097. *(24) Spong CY, Mercer BM, D'alton M, Kilpatrick S, Blackwell S, Saade G. Timing of indicated latepreterm and early-term birth. Obstet Gynecol 2011; 118(2 Pt 1):323-333.
EP
(25) MacKay DF, Smith GC, Dobbie R, Pell JP. Gestational age at delivery and special educational need: retrospective cohort study of 407,503 schoolchildren. PLoS Med 2010; 7(6):e1000289.
AC C
*(26) Monier I, Blondel B, Ego A, Kaminiski M, Goffinet F, Zeitlin J. Poor effectiveness of antenatal detection of fetal growth restriction and consequences for obstetric management and neonatal outcomes: a French national study. BJOG 2015; 122(4):518-527. (27) Duley L, Henderson-Smart DJ, Meher S, King JF. Antiplatelet agents for preventing preeclampsia and its complications. Cochrane Database Syst Rev 2007;(2):CD004659. (28) Bujold E, Roberge S, Lacasse Y, Bureau M, Audibert F, Marcoux S et al. Prevention of preeclampsia and intrauterine growth restriction with aspirin started in early pregnancy: a meta-analysis. Obstet Gynecol 2010; 116(2 Pt 1):402-414. (29) Dodd JM, McLeod A, Windrim RC, Kingdom J. Antithrombotic therapy for improving maternal or infant health outcomes in women considered at risk of placental dysfunction. Cochrane Database Syst Rev 2013; 7:CD006780.
19
ACCEPTED MANUSCRIPT *(30) Rodger MA, Hague WM, Kingdom J, Kahn SR, Karovitch A, Sermer M et al. Antepartum dalteparin versus no antepartum dalteparin for the prevention of pregnancy complications in pregnant women with thrombophilia (TIPPS): a multinational open-label randomised trial. Lancet 2014; 384(9955):1673-1683. (31) Lyall F. Development of the utero-placental circulation: the role of carbon monoxide and nitric oxide in trophoblast invasion and spiral artery transformation. Microsc Res Tech 2003; 60(4):402-411.
RI PT
(32) Schleussner E, Lehmann T, Kahler C, Schneider U, Schlembach D, Groten T. Impact of the nitric oxide-donor pentaerythrityl-tetranitrate on perinatal outcome in risk pregnancies: a prospective, randomized, double-blinded trial. J Perinat Med 2014; 42(4):507-514.
SC
(33) Abdel RM, El-Berry S, Abosereah M, Edris Y, Sharafeldeen A. Prophylactic treatment for preeclampsia in high-risk teenage primigravidae with nitric oxide donors: a pilot study. J Matern Fetal Neonatal Med 2016; 29(16):2617-2620.
M AN U
(34) Panda S, Das A, Md NH. Sildenafil citrate in fetal growth restriction. J Reprod Infertil 2014; 15(3):168-169. (35) Say L, Gulmezoglu AM, Hofmeyr GJ. Maternal oxygen administration for suspected impaired fetal growth. Cochrane Database Syst Rev 2003;(1):CD000137. (36) Spencer RN, Carr DJ, David AL. Treatment of poor placentation and the prevention of associated adverse outcomes--what does the future hold? Prenat Diagn 2014; 34(7):677684.
TE D
(37) Gulmezoglu AM, Crowther CA, Middleton P, Heatley E. Induction of labour for improving birth outcomes for women at or beyond term. Cochrane Database Syst Rev 2012; 6:CD004945.
EP
*(38) Wood S, Cooper S, Ross S. Does induction of labour increase the risk of caesarean section? A systematic review and meta-analysis of trials in women with intact membranes. BJOG 2014; 121(6):674-685.
AC C
(39) Walker KF, Bugg GJ, Macpherson M, McCormick C, Grace N, Wildsmith C et al. Randomized Trial of Labor Induction in Women 35 Years of Age or Older. N Engl J Med 2016; 374(9):813822. (40) Wood AM, Pasupathy D, Pell JP, Fleming M, Smith GC. Trends in socioeconomic inequalities in risk of sudden infant death syndrome, other causes of infant mortality, and stillbirth in Scotland: population based study. BMJ 2012; 344:e1552. (41) Smith GC. Researching new methods of screening for adverse pregnancy outcome: lessons from pre-eclampsia. PLoS Med 2012; 9(7):e1001274.
*Studies of particular importance to the current topic.
ACCEPTED MANUSCRIPT Table 1. Top 11 research priorities from a stillbirth priority setting partnership
TE D
M AN U
SC
RI PT
Research questions How can the structure and function of the placenta be assessed during pregnancy to detect potential problems and reduce the risk of stillbirth? Does ultrasound assessment of fetal growth in the third trimester reduce stillbirth? Do modifiable ‘lifestyle’ factors (e.g. diet, vitamin deficiency, sleep position, sleep apnea, lifting and bending) cause or contribute to stillbirth risk? Which investigations identify a fetus at risk of stillbirth after a mother believes she has experienced reduced fetal movements? Can the wider use of existing tests and monitoring procedures, especially in later pregnancy, and the development and implementation of novel tests (biomarkers) in the mother or in early pregnancy, help prevent stillbirth? What causes stillbirth in normally grown babies? What is the most appropriate bereavement and postnatal care for both parents following a stillbirth? Which antenatal care interventions are associated with a reduction in the number of stillbirths? Would more accessible evidence-based information on signs and symptoms of stillbirth risk, designed to empower women to raise concerns with healthcare professionals, reduce the incidence of stillbirth? How can staff support women and their partners in subsequent pregnancies, using a holistic approach to reduce anxiety, stress and any associated increased visits to healthcare settings? Why is the incidence of stillbirth in the UK higher than in other similar high-income countries, and what lessons can we learn from this?
AC C
EP
Text quoted verbatim from Heazell et al, 2015 (reference 3)
ACCEPTED MANUSCRIPT Table 1. Maternal risk factors for stillbirth. 95% Adjusted Odds ratio confidence interval Non-Hispanic black race/ethnicity 2.12 1.41 to 3.20 Previous stillbirth 5.91 3.18 to 11.00 Nulliparity + previous losses at <20 weeks' 3.13 2.06 to 4.75 Nulliparity, no previous losses 1.98 1.51 to 2.60 Diabetes mellitus 2.50 1.39 to 4.48 Maternal age 40 years or older 2.41 1.24 to 4.70 Maternal AB blood type 1.96 1.16 to 3.30 History of drug addiction 2.08 1.12 to 3.88 Smoking 1.55 1.02 to 2.35 Obesity/overweight 1.72 1.22 to 2.43 Not living with a partner 1.62 1.15 to 2.27 Multiple pregnancy 4.59 2.63 to 8.00
M AN U
SC
RI PT
Characteristic
AC C
EP
TE D
Data from reference 5 (see publication for details of multivariate analysis and referent categories). Stillbirth defined on basis of ≥20 weeks gestational age threshold.
ACCEPTED MANUSCRIPT Table 2. Maternal risk factors for gestational diabetes mellitus
AC C
EP
TE D
M AN U
SC
RI PT
Maternal risk factors for gestational diabetes mellitus Body mass index >30 kg/m2 Previous macrosomic infant (>4.5kg) Previous gestational diabetes First degree relative with diabetes Ethnic origin associated with high incidence of diabetes
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
ACCEPTED MANUSCRIPT Highlights
•
Stillbirth in the single major determinant of perinatal death
•
Screening for stillbirth currently targets investigations to women with risk factors, which are poorly discriminative There is scope for screening using ultrasound and biomarkers
•
The primary disease modifying intervention is delivery
•
Induction at term prevents perinatal death without increasing caesarean risk
AC C
EP
TE D
M AN U
SC
RI PT
•